pandas - groupby - substrack from first

import pandas as pd

df = pd.DataFrame({
    'b': ['s', 'j', 'm', 's', 'j'],
    'c': [12, 16, 5, 7, 8],
})
#print(df)

df['d'] = pd.np.zeros(df.shape[0])
#print(df)

for key, grp in df.groupby('b'):
    #print('---')
    #print(key)
    #print(grp)
    # it calculates first - sum(rest)
    result = grp['c'].iloc[0] - grp['c'].iloc[1:].sum()
    first_index = grp.index[0]
    df['d'][first_index] = result
    #print(key, result)

print(df)